Methods for recording and computer analysis of ballistocardiograms (Bcg) and percutaneous arterial pulse recordings (CP) have been developed by the applicant for use in large scale clinical trials. Analyses of Bcg and CP by these methods will be examined for effects of laboratory technique and possible regional differences in patients from Baltimore, Chicago, and Minneapolis laboratories of the National Coronary Drug Project (CDP). The many Bcg and CP measurements made by computer include time intervals in both systole and diastole, wave amplitudes and slopes, cross correlation coefficients for shape of pattern, and estimates of variability or rhythm effects. These variables are affected by laboratory technique to varying degrees, some very strongly, others moderately, and still others not at all. Advantage will be taken of these properties to identify effects of laboratory technique which affect poolability of data from the different sources adversely. Corrections will be applied where appropriate. Corrected variables, or variables without effects of technique, may reflect real differences in cardiac function of patients in the three locations. These steps for validation of the Bcg-CP data are essential for the long term goal of determination of prognostic power of the noninvasive cardiovascular methods. In addition, baseline recordings from 50 carefully selected normal males will be made using the newly developed techniques. Computer analyses of this group will provide normal standards for many of the new and unfamiliar variables which are not presently available for aid in interpretation of results from the CDP patients.